add object detection (introduction)

Author: yiyujin

docs/reference/object-detection.md

@@ -3,4 +3,37 @@
 <center>
   <img class="header-img" src="assets/header-image-classifier.png" alt="ImageClassifier Header Image" >
   <p class="img-credit"> Image Credit: <a href="https://thenounproject.com/creator/naveena.160" target="_blank" title="Naveen">Naveen</a> | <a href='mailto:[email protected]'>Contribute ♥️</a> </p>
-</center>
+</center>
+
+## Description
+
+The ml5.js objectDetection is a pre-trained model that can detect object from an image or a video.
+
+### Supported models
+The ml5.js objectDetection uses pre-trained **Tensorflow.js CocoSsd model** by default.
+
+Support for more models such as Transformer.js is under development.
+
+### Model and Data Provenance
+ml5.js team believes that it is important to not only use the model, but to understand where it is coming from : what data was used, who trained and for waht reason. For more information on the model, please visit [Model and Data Provenance](https://archive-docs.ml5js.org/#/reference/object-detector?id=model-and-data-provenance)
+
+
+### Functionality
+The ml5.js objectDetection detects objects in a source input.
+
+The source input can be :
+
+- **Webcam**
+- **Image File**
+- **Video File**
+
+## Quick Start
+Run and explore pre-built examples!
+
+### Examples - p5 sketches
+- [ObjectDetection Webcam]()
+- [ObjectDetection Image]()
+- [ObjectDetection Video]()
+
+### Video Tutorials
+- [ml5.js: Object Detection with COCO-SSD](https://youtu.be/QEzRxnuaZCk) by the Coding Train